Lithium ion battery is a prime candidate energy storage device for electric vehicle (EV), renewable energy storage, and smart grid applications. Graphite materials have been widely used as an anode active material for commercial lithium ion batteries due to their relatively low cost and excellent reversibility. However, the theoretical lithium storage capacity of graphite is only 372 mAh/g (based on LiC6), which can limit the total capacity and energy density of a battery cell. The emerging EV and renewable energy industries demand the availability of rechargeable batteries with a significantly higher energy density and power density than what the current Li ion battery technology can provide. Hence, this requirement has triggered considerable research efforts on the development of electrode materials with higher specific capacity, excellent rate capability, and good cycle stability for lithium ion batteries.
Several elements from Group HI, IV, and V in the periodic table can form alloys with Li at certain desired voltages. Therefore, various anode materials based on such elements and some metal oxides (e.g., SnO2) have been proposed for lithium ion batteries. Among these, silicon is considered the most promising one since it has the highest theoretical specific capacity (up to 4,200 mAh/g in the stoichiometric form of Li4.4Si) and low discharge potential (i.e., high operation potential when paired with a cathode). However, the dramatic volume change (up to 380%) of Si during lithium ion alloying and de-alloying (cell charge and discharge) often leads to severe and rapid battery performance deterioration. The performance fade is mainly due to the volume change-induced pulverization of Si and the inability of the binder/conductive additive to maintain the electrical contact between the pulverized Si particles and the current collector. In addition, the intrinsic low electric conductivity of silicon is another challenge that needs to be addressed. Thus far, many attempts have been made to improve the electrochemical performance of Si-based anode materials, which include (1) reducing particle size to the nano-scale (<100 nm), such as Si nanoparticles, nanowires, or thin film, to reduce the total strain energy, which is a driving force for crack formation in the particle; (2) depositing Si particles on a highly electron-conducting substrate; (3) dispersing Si particles in an active or non-active matrix; and (4) coating Si particles with a layer of carbon. Although some promising anodes with specific capacities in excess of 1,000 mAh/g have been reported, it remains challenging to retain such high capacities over cycling (e.g., for more than 100 cycles) without significant capacity fading.
Our research group discovered graphene, a new class of nano carbon materials, in 2002 [B. Z. Jang, et al, “Nano-scaled Graphene Plates,” U.S. patent application Ser. No. 10/274,473 (Oct. 21, 2002); now U.S. Pat. No. 7,071,258 (Jul. 4, 2006)]. A single-layer graphene is composed of carbon atoms forming a 2-D hexagonal lattice through strong in-plane covalent bonds. In a multi-layer graphene, several graphene planes are wealdy bonded together through van der Waals forces in the thickness-direction. Since 2002, our research group has been working to develop processes for mass-producing both single-layer and multi-layer graphene and their composites, and to investigate the applications of graphene materials in many areas. Recently, we have demonstrated that graphene sheets can serve as a supporting substrate for an anode active material, which can be a thin film coated onto a graphene surface or fine powders (e.g. nanoparticles) bonded to a graphene surface [B. Z. Jang and A. Zhuma, “Nano Graphene Platelet-Based Composite Anode Compositions for Lithium Ion Batteries,” U.S. patent application Ser. No. 11/982,672 (Nov. 5, 2007) now U.S. Pat. No. 7,745,047 (Jun. 29, 2010)]. Several other research groups have also reported a similar approach of combining graphene with an anode active material (e.g., SnO2, TiO2, Mn3O4, Fe2O3, and Co3O4), by taking advantages of graphene's electric conductivity.
We hypothesize that, although the Si particles in a graphene-silicon hybrid material can still expand and shrink during lithiation/delithiation (Li alloying/de-alloying during cell charge/discharge), the strong but flexible graphene sheets surrounding the Si nanoparticles are capable of cushioning the stress/strain to some extent. Meanwhile, the graphene sheets ensure good electric contacts between adjacent Si particles and between Si particles and a current collector. With an ultra-high length-to-thickness aspect ratio (up to 50,000) and low thickness (e.g. just one or a few atomic layers), a very small amount of graphene is sufficient to provide the electron-conducting network. Furthermore, graphene by itself could also be a good anode active material. Therefore, graphene-silicon composites have been proposed as promising anode materials for lithium ion batteries.
However, such graphene/silicon composite anode materials were made by simply mixing silicon nanoparticles with graphene or graphene oxide, and such an approach often led to only limited improvement in electrochemical performance. In contrast to graphene-metal oxide composite anode materials in which metal oxide can be readily deposited or grown on graphene sheet surfaces, the limited success in graphene-silicon composite is due to the lack of a simple and efficient method capable of well controlling the structure and morphology of the resulting hybrid material that could deliver the aforementioned features and advantages.
Furthermore, from mass production and cost perspectives, current processes for producing nano Si powder have been time-consuming, energy-intensive, requiring the use of high-vacuum, high-temperature, and/or high-pressure production equipment. The resulting Si nano powder products have been extremely expensive and this cost issue has severely impeded the full-scale commercialization of Si nano powder materials. Hence, there exists a strong need for a more cost-effective process for producing Si nano powder in large quantities.
More significantly, current processes for producing Si/graphene hybrid materials typically entail producing Si nano particles and graphene sheets separately and then combining the two components together. Such a simple-minded approach eliminates the possibility for graphene or graphene precursor to offer the beneficial effects on the Si nano particle forming kinetics and energetics, and on the structure and morphology of the resulting hybrid materials. This is beyond and above the simple issues of the higher costs associated with the conventional processes. The present invention addresses all of these longstanding and most challenging problems in the lithium-ion battery industry.
The past research and development efforts on silicon nano materials have been focused mostly on silicon nano particles, silicon nano wires, silicon thin films, and even silicon nano tubes. Common methods used for producing silicon nano powders include plasma-enhanced chemical vapor deposition (PECVD), laser-induced pyrolysis of SiH4, and hot-wire synthesis methods. These techniques require either ultra-high temperature or high power supply, and sometimes ultra-high vacuum, which lead to high fabrication costs. The following references are related to these conventional processes:    1. M. R. Scriba, C. Arendse, M. Härting, D. T. Britton, Thin Solid Films 516 (2008) 844-846.    2. Mark T. Swihart, et al., Process for producing luminescent silicon nanoparticles, U.S. Pat. No. 7,371,666 (May 13, 2008).    3. Ming Li, et al., Single wafer thermal CVD processes for hemispherical grained silicon and nano-crystalline grain-sized polysilicon, U.S. Pat. No. 7,341,907 (Mar. 11, 2008).    4. Byoung-lyong Choi, et al., Silicon nano wires, semiconductor device including the same, and method of manufacturing the silicon nano wires, U.S. Pat. No. 7,625,812 (Dec. 1, 2009).    5. Chi-Pin Lu, et al., Method of manufacturing nano-crystalline silicon dot layer, U.S. Pat. No. 7,927,660 (Apr. 19, 2011).    6. Sushil Kumar, et al., Process for the preparation of photo luminescent nanostructured silicon thin films, US Pub. No. 2010/0285235 (Nov. 11, 2010).    7. L. W. Wu, et al., Nano powder production system, US Pub. No. 2003/0108459 (Jun. 12, 2003).    8. Hai-Lin Sun, et al., Method of manufacturing silicon nano-structure, U.S. Pat. No. 7,888,271 (Feb. 15, 2011).    9. Shinichiro Ishihara, et al., Method for producing an amorphous silicon semiconductor device using a multichamber PECVD apparatus, U.S. Pat. No. 4,800,174 (Jan. 24, 1989).    10. Z. H. Bao, M. R. Weatherspoon, S. Shian, et al., “Chemical reduction of three-dimensional silica micro-assemblies into microporous silicon replicas,” Nature, 446 (2007) 172-175.    11. E. K. Richman, C. B. Kang, T. Brezesinski and S. H. Tolbert, “Ordered meso-porous silicon through magnesium reduction of polymer template silica thin films,” Nano Letters, 8 (2008) 3075-3079.
A promising method for reprocessing silica via magnesiothermic reduction was proposed by Bao, et al. [Ref. 10] using magnesium vapor to produce meso-porous silicon. Then, Richman et al. reported a similar study of reducing silica thin film to meso-porous silicon thin film by magnesiothermic routine [Ref. 11]. Silicon is obtained by the following reaction with Mg: 2Mg+SiO2→2 MgO+Si.
Magnesiothermic reduction of silica requires much lower temperatures (normally in the range of 600-800° C.) compared with the carbothermal reduction of silica (normally over 2000° C.) and thus has become a popular technique used in pure metal production. The patent publications given below provide some methods utilizing magnesiothermic reduction to produce pure metal or silicon.
U.S. Pat. No. 7,615, 206 issued in 2009 to K. H. Sandhage and Z. H. Bao refers to methods for the production of shaped nanoscale-to-microscale silicon through partially or completely converting a nanoscale-to-microscal silica template by using magnesium vapor. After the reduction of silica to silicon, the original shape could be well maintained.
US Patent Publication No. US 2010/0288649 (inventor: U. B. Pal) provides a process and apparatus that allow metals, including metals having stable oxide phases and metals with variable valences, to be extracted from their respective ores via a reducing chamber by highly reactive metal (e.g. Mg). A solid oxide membrane (SOM) process is used to generate vapor of the highly reactive metal (Mg) in the electrolysis chamber.
US Patent Publication No. US 2010/0092141 (inventors: G. F. Li and F. Yaman) disclosed a method of converting silica to silicon and fabricating silicon photonic crystal fiber (PCF) using basically magnesiothermic reduction in a sealed or unsealed container.
US Patent Publication No. US 2011/0085960 (inventors: A. Mukasyan, et al.) disclosed a method for synthesis of high surface area (>100 m2/g) and nano-sized (50-200 nm) silicon powder by initiation of self-sustained combustion reaction in a mixture of silicon dioxide and magnesium powders in a sealed reactor chamber under pressurized inert gas atmosphere. A specific feature of the method is rapid cooling of the product at a rate of 100 K/s to 400 K/s in the area directly behind the combustion front.
U.S. Pat. No. 7,972,584 issued to J. G. Blencoe in 2011 provides the magnesiothermic methods of producing solid silicon using magnesium gas having a purity of from 98.0 to 99.999%.
All these methods employ magnesiothermic reduction to form silicon from its precursor-silica either by Mg vapor or by Mg powder. When using Mg vapor to reduce silica, magnesium silicide could be easily formed and, hence, this process is not suitable for mass production. Using magnesium powder will add to cost of producing nano-sized silicon and the particle size of magnesium could dramatically influence the reduction results and purity, and thus is not suitable for mass production.
Herein, we present a facile and cost-effective method of mass-producing silicon nano powder and graphene-doped silicon nano powder. This method entails mixing a graphene material (e.g. pristine graphene, graphene oxide, or graphene fluoride) with nano-sized silica (or other silicon precursor) to obtain graphene-supported nano-sized silica and then reducing the graphene-supported nano-sized silicon via the magnesiothermic method to obtain graphene-doped nano-Si. The resulting nano-sized silicon has a size typically in the range of 2 nm-50 nm (more often in the range of 5-40 nm, and most often 10-20 nm). The Si nano particles can be prepared in the size range of 50-100 nm if so desired. Pure nano silicon powders could be obtained by heat treating the graphene-doped nano Si powders to remove graphene in the temperature range of 400˜800° C.